-
Notifications
You must be signed in to change notification settings - Fork 167
/
run-egogesture.sh
executable file
·75 lines (66 loc) · 1.99 KB
/
run-egogesture.sh
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
#"$1" pretrain path
#"$2" model
#"$3" width_mult
python main.py --root_path ~/ \
--video_path ~/datasets/EgoGesture \
--annotation_path Real-time-GesRec-2/annotation_EgoGesture/egogestureall_but_None.json \
--result_path Real-time-GesRec-2/results \
--pretrain_path "$1" \
--dataset egogesture \
--n_classes 27 \
--n_finetune_classes 83 \
--model "$2" \
--width_mult "$3" \
--model_depth 101 \
--resnet_shortcut B \
--resnext_cardinality 16 \
--train_crop random \
--learning_rate 0.01 \
--sample_duration 32 \
--modality Depth \
--pretrain_modality RGB \
--downsample 1 \
--batch_size 24 \
--n_threads 16 \
--checkpoint 1 \
--n_val_samples 1 \
--n_epochs 60 \
--ft_portion complete \
# --no_train \
# --no_val \
# --test
# --resume_path Efficient-3DCNNs/results/egogesture_shufflenet_2.0x_Depth_16_checkpoint.pth \
# --pretrain_path Efficient-3DCNNs/results/jester_mobilenetv2_1.0x_RGB_16_best.pth \
# --pretrain_path Efficient-3DCNNs-Ahmet/results/jester_resnext_101_RGB_16_checkpoint.pth \
# --n_finetune_classes 83 \
# --test_subset val \
# python main.py --root_path ~/ \
# --video_path ~/datasets/EgoGesture \
# --annotation_path Efficient-3DCNNs-Ahmet/annotation_EgoGesture/egogestureall_but_None.json \
# --result_path Efficient-3DCNNs-Ahmet/results \
# --pretrain_path "$1" \
# --dataset egogesture \
# --n_classes 27 \
# --n_finetune_classes 84 \
# --model "$2" \
# --groups 3 \
# --version 1.1 \
# --width_mult "$3" \
# --train_crop random \
# --learning_rate 0.1 \
# --sample_duration 16 \
# --modality RGB \
# --downsample 2 \
# --batch_size 80 \
# --n_threads 16 \
# --checkpoint 1 \
# --n_val_samples 1 \
# --n_epochs 20 \
# --train_validate \
# --test_subset test \
# --ft_portion last_layer
# # --no_train \
# # --no_val \
# # --test
# # --resume_path Efficient-3DCNNs/results/egogesture_shufflenet_2.0x_Depth_16_checkpoint.pth \
# # --pretrain_path Efficient-3DCNNs/results/jester_mobilenetv2_1.0x_RGB_16_best.pth \